ultrix 4.0 stability

[Steve Simmons]

emv at math.lsa.umich.edu (Edward Vielmetti) writes:

>if there's someone who's running ultrix 4.0 & would like to
>comment on its stability I'm interested.  just got in 3
>ds5000's with some tapes that say ultrix 4.0 plus a tape that
>says "mandatory patches" with the promise of more patches to 
>come.

>especially looking for installation gotchas....

I spoke to Ed face-to-face the other day, but will repeat for
those generally interested:

We had a beta copy of 4.0 until yesterday on a 3100 (my own workstation).
I beat it, used DecWindows, plain X, X w/uwm, cross mounted NFS stuff to
3.1, 3.1C, Suns, and a Gould.  In general, it worked.  I found one and
only one bug (in su), which was susceptable to an easy workaround.  It's
been reported to DEC, and was hopefully fixed before release.

I ran 3.1 and 3.1C binaries on 4.0beta with no problems, but did not
try the reverse.

It did not have ndbm, and still ran BSD 4.2 syslog rather than 4.3.  I did
did not try building either one on Ultrix 4.0, but note that they were
largely compile and go if you have BSD source.  Sorry, I will not
distribute binaries for the unlicenced.

It will run the IDA sendmail 5.59 available by anon ftp from
gatekeeper.dec.com.

We did not run DEC bind.